2.6.26-3 oopses on boot on Sunblade 1000

2008-08-24 Thread Jurij Smakov
Hi,

I've installed the new kernel today (2.6.26-1-sparc64-smp, version 
2.6.26-3) and it oopses during boot on my Sunblade1000:

[  131.424082] qla2xxx 0001:00:04.0: Allocated (252 KB) for firmware dump...
[  131.764012] qla2xxx 0001:00:04.0: LIP reset occured (f7e8).
[  131.948876] scsi0 : qla2xxx
[  131.963025] qla2xxx 0001:00:04.0: LIP occured (f7e8).
[  131.966458] qla2xxx 0001:00:04.0: LOOP UP detected (1 Gbps).
[  131.967797] Unable to handle kernel NULL pointer dereference
[  131.967808] tsk-{mm,active_mm}-context = 01cc
[  131.967815] tsk-{mm,active_mm}-pgd = f8007bfcc000
[  131.967824]   \|/  \|/
[  131.967829]   @'/ .. \`@
[  131.967833]   /_| \__/ |_\
[  131.967837]  \__U_/
[  131.967843] qla2xxx_0_dpc(807): Oops [#1]
[  131.967855] TSTATE: 004480009604 TPC: 1009c7dc TNPC: 
1009c7e0 Y: 031bNot tainted
[  131.967904] TPC: fc_flush_work+0x1c/0x40 [scsi_transport_fc]
[  131.967914] g0: 0075c400 g1:  g2: f8007b004000 
g3: 000b00b6
[  131.967924] g4: f8007b021780 g5: f8007ef34000 g6: f8007b054000 
g7: 2000870bf4b6
[  131.967934] o0: 100a1590 o1: 0003 o2: 0001 
o3: 
[  131.967944] o4:  o5: 7fffe000 sp: f8007b0570c1 
ret_pc: 0045033c
[  131.967966] RPC: complete+0x44/0x48
[  131.967974] l0: f8007b006960 l1: 000f l2: f8007e4000b0 
l3: 
[  131.967984] l4: f8007e048070 l5: f8007e006200 l6: 00d0 
l7: 0020
[  131.967994] i0: f8007b004000 i1:  i2:  
i3: f8007e0b4b60
[  131.968004] i4:  i5:  i6: f8007b057181 
i7: 1009edb8
[  131.968033] I7: fc_remote_port_add+0x18/0x664 [scsi_transport_fc]
[  131.968056] Caller[1009edb8]: fc_remote_port_add+0x18/0x664 
[scsi_transport_fc]
[  131.968080] Caller[100b3ff8]: qla2x00_update_fcport+0x2b8/0x360 
[qla2xxx]
[  131.968164] Caller[100b499c]: qla2x00_configure_loop+0x85c/0x1a18 
[qla2xxx]
[  131.968194] Caller[100b5bd4]: qla2x00_loop_resync+0x7c/0x10c 
[qla2xxx]
[  131.968221] Caller[100b1f7c]: qla2x00_do_dpc+0x60c/0x6e8 [qla2xxx]
[  131.968248] Caller[0046e378]: kthread+0x4c/0x78
[  131.968263] Caller[00426df8]: kernel_thread+0x38/0x48
[  131.968275] Caller[0046e1a0]: kthreadd+0xc4/0x1a0
[  131.968285] Instruction DUMP: 11040285  84100018  90122190 c2586330 
0ac84007  b011  c258a0b0  7c0ef9ee  d2586008 
[  139.589175] qla2xxx 0001:00:04.0: 
[  139.589179]  QLogic Fibre Channel HBA Driver: 8.02.01-k4
[  139.589184]   QLogic QLA22xx - 
[  139.589188]   ISP2200: PCI (66 MHz) @ 0001:00:04.0 hdma-, host#=0, 
fw=2.02.08 TP

Does anyone see anything similar? If you have access to a machine with 
QLA2 SCSI controller, can you please try installing the latest 
kernel from unstable and making sure that it works for you?

Thanks.
-- 
Jurij Smakov   [EMAIL PROTECTED]
Key: http://www.wooyd.org/pgpkey/  KeyID: C99E03CC


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Re: 2.6.26-3 oopses on boot on Sunblade 1000

2008-08-24 Thread Jurij Smakov
On Sun, Aug 24, 2008 at 11:55:08AM +0100, Jurij Smakov wrote:
 Hi,
 
 I've installed the new kernel today (2.6.26-1-sparc64-smp, version 
 2.6.26-3) and it oopses during boot on my Sunblade1000:

It is not 100% reproducible, so far it triggered in 2 out of 6 boot 
attempts.

Cheers.
-- 
Jurij Smakov   [EMAIL PROTECTED]
Key: http://www.wooyd.org/pgpkey/  KeyID: C99E03CC


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



New SILO on its way to unstable

2008-08-24 Thread Jurij Smakov
Hi,

I've just uploaded new SILO (1.4.13a+git20070930-3) to unstable to fix 
the RC bug #495860. If you are running unstable, please upgrade to 
this version in the next couple of days and report any problems.

Thanks.
-- 
Jurij Smakov   [EMAIL PROTECTED]
Key: http://www.wooyd.org/pgpkey/  KeyID: C99E03CC


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



PCI cards on Ultra5

2008-08-24 Thread Christian Euler
Hi everybody,

I plugged a PCI USB card in my good old Ultra 5 box (running Debian Etch
with 2.6.24 kernel), but the card doesnt show up in lspci.

Is there anything special about PCI support on UltraSparc? 

There are some jumpers located on the riser card, but since I dont have
any documentation I dont want to play around with them :)

Regards,
Chris



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Re: PCI cards on Ultra5

2008-08-24 Thread Geert Stappers
Op 20080824 om 15:25 schreef Christian Euler:
 Hi everybody,
 
 I plugged a PCI USB card in my good old Ultra 5 box (running Debian Etch
 with 2.6.24 kernel), but the card doesnt show up in lspci.
 
 Is there anything special about PCI support on UltraSparc? 

Not as far I known.

I think the original poster is encountering hardware defects.

Hardware test to preform:
* place a known to be working PCI (Network) card in the Ultra 5

* place the PCI USB card in known to be working PCI slot.


HtH
G


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Re: PCI cards on Ultra5

2008-08-24 Thread Christian Euler
Hi Geert,

  Is there anything special about PCI support on UltraSparc?
 Not as far I known.

 I think the original poster is encountering hardware defects.

 Hardware test to preform:
 * place a known to be working PCI (Network) card in the Ultra 5

 * place the PCI USB card in known to be working PCI slot.

I ripped the USB card off my wife's PC (where it was working fine).
I also tried other cards on the Ultra, without success.

The machine itself runs pretty stable, I used it for about 8 hours yesterday 
without any glitches. So I don't want to talk about a hardware problem yet. ;)

Is there any possibility to debug this (from PROM monitor for example) ?

Btw, here's the lspci output (if that helps) :

rincewind:~# lspci
00:00.0 Host bridge: Sun Microsystems Computer Corp. Ultra IIi
00:01.0 PCI bridge: Sun Microsystems Computer Corp. Simba Advanced PCI Bridge 
(rev 13)
00:01.1 PCI bridge: Sun Microsystems Computer Corp. Simba Advanced PCI Bridge 
(rev 13)
01:01.0 Bridge: Sun Microsystems Computer Corp. EBUS (rev 01)
01:01.1 Ethernet controller: Sun Microsystems Computer Corp. Happy Meal (rev 01)
01:02.0 VGA compatible controller: ATI Technologies Inc 3D Rage Pro 215GP (rev 
5c)
01:03.0 IDE interface: Silicon Image, Inc. PCI0646 (rev 03)

Regards,
Chris


signature.asc
Description: This is a digitally signed message part.


Re: PCI cards on Ultra5

2008-08-24 Thread Admar Schoonen
Hi,

On Sun, Aug 24, 2008 at 07:59:19PM +0200, Christian Euler wrote:
 Hi Geert,
 
   Is there anything special about PCI support on UltraSparc?
  Not as far I known.
 
  I think the original poster is encountering hardware defects.
 
  Hardware test to preform:
  * place a known to be working PCI (Network) card in the Ultra 5
 
  * place the PCI USB card in known to be working PCI slot.
 
 I ripped the USB card off my wife's PC (where it was working fine).
 I also tried other cards on the Ultra, without success.
 
 The machine itself runs pretty stable, I used it for about 8 hours yesterday
 without any glitches. So I don't want to talk about a hardware problem yet. ;)

There are various success and fail stories of putting USB PCI cards in Ultra
5's. I once heard that U5's only like PCI 2.1 cards and not 2.2 (see e.g.
http://lists.debian.org/debian-sparc/2007/12/msg00012.html). Don't know if it's
true.

You might try a different USB card.

HTH

Admar


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



[Solved] Re: PCI cards on Ultra5

2008-08-24 Thread Christian Euler
Hi,

I got it working :)

There is a small jumper pad on the riser card (I mentioned it in my original 
post) labeled JSCAN RING END POINT
with one jumper for each PCI slot and possible settings PASS and END.

In my case, all jumpers were set to PASS. I put the jumper for PCI slot 3 to 
END and now my box recognises the USB card.

Sun hardware is really weird. Anyway, thanks for all your suggestions ;)

Regards,
Chris



signature.asc
Description: This is a digitally signed message part.


[Fwd: Re: Shorewall and SPARC]

2008-08-24 Thread Eric Nichols
I apologize for the long thread.  This was a discussion I had with the
shorewall maintainer for debian.  Might be a sparc issue?

-- Original Message --
Subject: Re: Shorewall and SPARC
From:Roberto C. Sánchez [EMAIL PROTECTED]
Date:Thu, August 21, 2008 8:05 pm
To:  Eric Nichols [EMAIL PROTECTED]
--

On Thu, Aug 21, 2008 at 07:25:50PM -0400, Eric Nichols wrote:
 Roberto C. Sánchez wrote:
 On Thu, Aug 21, 2008 at 12:34:57PM -0400, Eric Nichols wrote:
 Hello,
 I've been using shorewall in Etch for quite a while and I absolutely
 love it.  I am currently running this on a Sun Ultra 2 (SPARC processor)
 and get the following messages below.  Should I be concerned about them?
  I can't find anything that references these items.
 
 ip_tables: conntrack match: invalid size 80 != 72
 ip_tables: connmark match: invalid size 24 != 16
 ip_tables: MARK target: invalid size 16 != 8
 ip_tables: CONNMARK target: invalid size 24 != 16
 
 OK.  I should have been able to diagnose this right away, but it escaped
 me.  After consultation with the author of Shorewall, I can point you at
 this: http://www.shorewall.net/3.0/FAQ.htm#faq61
 
 Basically, I think that this can be caused by a partially upgraded
 system.  Did you happen to upgrade iptables but not the kernel?
 
 Regards,
 
 -Roberto
 
 I've seen this on just about every sparc install I've done.  I'm running
 stock etch with everything current.  I'm leaning more towards an issue
 in the sparc build itself.  It's not the most widely maintained
 architecture.  Thanks for the followup.

I've never used Debian on Sparc hardware, so I don't know.  However,
according to Tom (the author of Shorewall), this is a common occurrence
on Debian and it has to do with iptables being built against an
incompatible set of kernel headers (as compared to the kernel against
which it is running).

If you try asking your question on the mailing list (and at least
mention that you have already been give an answer, but that you are
looking for more info), you might find a Debian+Sparc user who can help.
You might also try the netfilter mailing list, as this is not
Shorewall-specific.

Regards,

-Roberto

-- 
Roberto C. Sánchez
http://people.connexer.com/~roberto
http://www.connexer.com



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]